BA, gemcitabine, 5-aminoimidazole-4-carboxamide 1-β-D-ribofuranoside (AICAR), dimethyl sulfoxide (DMSO) and MTT were purchased from Sigma-Aldrich; Merck KGaA (Darmstadt, Germany). BA and AICAR were initially dissolved in dimethyl sulfoxide at stock concentrations of 50 mM and 2 M, respectively. Working concentrations for BA and AICAR were diluted immediately in culture medium prior to use. Human epidermal growth factor (EGF) and fibroblast growth factor (FGF) were purchased from PeproTech, Inc. (Rocky Hill, NJ, USA).The antibodies used in this study were as follows: Rabbit anti-Sox2 (1:1,000 dilution; cat. no. ab97959), anti-Oct4 (1:1,000 dilution; cat. no. ab18976) and anti-Nanog (1:1,000 dilution; cat. no. ab80892) (all from Abcam, Cambridge, UK), mouse anti-β-actin (1:10,000 dilution; cat. no. 19526; KangChen Bio-Tech, Inc., Shanghai, China), mouse anti-E-cadherin (1:1,000 dilution; cat. no. sc-71008), rabbit anti-cluster of differentiation (CD)133 (1:1,000 dilution; cat. no. sc-11406), mouse anti-aldehyde dehydrogenase (ALDH)1 (1:1,000 dilution; cat. no. sc-374149) and mouse anti-epithelial cell adhesion molecule (EpCAM) (1:1,000 dilution; cat. no. sc-66020) (all from Santa Cruz Biotechnology, Inc., Dallas, TX, USA), and rabbit anti-vimentin (1:800 dilution; cat. no. 12826), anti-AMPK (1:800 dilution; cat. no. 5831) and anti-phosphorylated (P)-AMPK (Thr172) (1:800 dilution; cat. no. 2535) (all from Cell Signaling Technology, Inc., Danvers, MA, USA).

The Mia PaCa-2 and Panc-1 human pancreatic cancer cell lines were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). Mia PaCa-2 and Panc-1 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (HyClone; GE Healthcare Life Sciences, Logan, UT, USA), 100 U/ml penicillin and 100 µg/ml streptomycin (Gibco; Thermo Fisher Scientific, Inc.). All cells were cultured in a humidified atmosphere at 37°C containing 5% CO₂.

Mia PaCa-2 and Panc-1 cells were seeded into 96-well plates at a density of 5×10³ cells/well and treated with various concentrations of BA (0, 12.5, 25, 50, 100 and 200 µM) for 24, 48 and 72 h at 37°C. Subsequently, 10 µl MTT (Sigma-Aldrich; Merck KGaA) was added and incubated for 4 h at 37°C. The supernatant was then replaced with 100 µl DMSO (Sigma-Aldrich; Merck KGaA) and absorbance was detected at 490 nm using a multiwell microplate reader (BioTek Instruments, Inc., Winooski, VT, USA).

Mia PaCa-2 and Panc-1 cells were digested, counted and seeded into 6-well plates at a density of 800 cells/well. After adherence overnight, pancreatic cancer cells were treated with BA (50 µM), gemcitabine (5 µM), or gemcitabine (5 µM) combined with BA (50 µM) for 24 h at 37°C, after which the medium was replaced with drug-free medium containing 10% FBS. The cells were then allowed to grow for 2 weeks to form colonies. After 2 weeks, the colonies were washed with PBS, fixed with 4% paraformaldehyde for 30 min at room temperature and stained for 10 min at room temperature with 0.1% crystal violet solution, followed by rinsing and imaging. The number of colonies with a diameter >0.5 mm was counted under a microscope (Nikon Eclipse Ti-S; Nikon Corporation, Tokyo, Japan).

The apoptosis of pancreatic cancer cells was assessed using flow cytometry with an Annexin V-fluorescein isothiocyanate (FITC)/7-aminoactinomycin D (7-AAD) apoptosis detection kit (BD Biosciences, Franklin Lakes, NJ, USA), according to the manufacturer’s protocol. The preprocessing of pancreatic cancer cells was performed as previously described (36). Cells treated with gemcitabine (5 µM), BA (50 µM), or gemcitabine (5 µM) combined with BA (50 µM) for 24 h at 37°C were washed twice in cold PBS and resuspended in 1X binding buffer (BD Biosciences) at a density of 1×10⁶ cells/ml. Cell suspensions (2.5×10⁵ cells) were added to 1.5 ml Eppendorf tubes, to which 5 µl allophycocyanin Annexin V and 7-AAD were added, and were gently vortexed. Subsequently, cells were incubated at room temperature for 15 min in the dark, and the percentage of apoptotic cells was quantified by flow cytometry using a FACSCalibur flow cytometer (BD Biosciences). Data were analyzed using Winmdi2.9 software (The Scripps Research Institute, San Diego, CA, USA). The total apoptotic rate was assessed by adding the rate of Annexin V-FITC⁺/7-AAD^- cells (early apoptotic cells) and Annexin V-FITC⁺/7-AAD⁺ cells (late apoptotic cells) together.

Following treatment with BA (50 µM) or AICAR (2 mM), or with BA (50 µM) combined with siRNA to knockdown AMPK or siControl for 24 h at 37°C, pancreatic cancer cells were digested, counted and seeded in 6-well ultra-low attachment plates (Corning Incorporated, Corning, NY, USA) at a density of 5,000 cells/well in serum-free DMEM/F12 medium (Gibco; Thermo Fisher Scientific, Inc.) containing 20 ng/ml human FGF, 20 ng/ml human EGF and 1% B27 (Gibco; Thermo Fisher Scientific, Inc.). Cells were subsequently cultured at 37°C in an atmosphere containing 5% CO₂ for 1 week to form tumorspheres. After 1 week, tumorsphere formation was counted and recorded under a light microscope (Nikon Corporation) at a magnification of ×200.

The wound-scratch assay was conducted to examine the migratory capacity of pancreatic cancer cells. Briefly, once Mia PaCa-2 and Panc-1 cells were cultured to 90-100% confluence, a 10-µl pipette tip was used to generate a wound in the surface of the cells in a 6-well plate, and then the cells were treated with or without BA (50 µM) for 48 h at 37°C. Images of the same fields at the indicated time-points (0 and 48 h) were captured under a light microscope (Nikon Corporation) at a magnification of ×100.

Transwell chamber (EMD Millipore, Billerica, MA, USA) assays were performed to assess the invasive ability of pancreatic cancer cells, in accordance with a previously described protocol (37). Briefly, Mia PaCa-2 and Panc-1 cells were serum-starved for 6-8 h and were pretreated with BA (50 µM) or AICAR (2 mM) for 24 h at 37°C. In addition, the upper chamber was coated with Matrigel (Sigma-Aldrich; Merck KGaA) and was incubated at 37°C in an atmosphere containing 5% CO₂ for 5 h. Subsequently, Mia PaCa-2 and Panc-1 cells (1×10⁵) were digested, resuspended in serum-free medium and seeded into the upper chamber. The pancreatic cancer cells were allowed to invade into the lower chamber, which contained medium supplemented with 10% FBS, for 24 h at 37°C. The non-invading cells on the upper side were scraped off using a cotton swab, and the membrane was then fixed with 4% paraformaldehyde for 30 min at room temperature and stained for 10 min at room temperature with 0.1% crystal violet. Subsequently, the number of cells on each membrane was counted in 10 random fields and images were captured under a light microscope (Nikon Corporation) at ×200 magnification. The values reported are the mean of triplicate experiments.

Following treatment with or without BA (50 µM), with various concentrations of gemcitabine (0, 1 and 5 µM), or with gemcitabine (5 µM) combined with BA (50 µM) for 24 h at 37°C, total RNA was extracted from the cells using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer’s protocol. The isolated total RNA was then reverse transcribed into cDNA using a PrimeScript RT reagent kit (Takara Biotechnology Co., Ltd., Dalian, China), according to the manufacturer’s protocol. An iQ5 Multicolor Real-Time PCR Detection system (Bio-Rad Laboratories, Inc., Hercules, CA, USA) and a SYBR Green PCR kit (Takara Biotechnology Co., Ltd.) were used to conduct RT-qPCR, according to the manufacturers’ protocols. The RT-qPCR experimental steps were conducted as previously described (38). The specificity of the amplified PCR products was evaluated by melting curve analysis, and the comparative Cq method was used to determine the expression levels of each target gene, with GAPDH as a normalization control, as previously described (39). The primer sequences used for RT-qPCR were as follows: Sox-2, forward 5′-GCCGAGTGGAAACTTTTGTCG-3′, reverse 5′-GGCAGCGTGTACTTATCCTTCT-3′; Nanog, forward 5′-TTTGTGGGCCTGAATAAGCAG-3′, reverse 5′-AGGGCTGTCCTGAATAAGCAG-3′; Oct4, forward 5′-CTGGGTTGATCCTCGGACCT-3′, reverse 5′-CCATCGGAGTTGCTCTCCA-3′; CD133, forward 5′-TACAACGCCAAACCACGACTGT-3′, reverse 5′-TCTGAACCAATGGAATTCAAGACCCTTT-3′; ALDH1, forward 5′-TGTTAGCTGATGCCGACTTG-3′, reverse 5′-CTTCTTAGCCCGCTCAACAC-3′; EpCAM, forward 5′-ATGTTTGGTGATGAAGGCAGAA-3′, reverse 5′-ATCGCAGTCAGGATCATAAAGC-3′; and β-actin, forward 5′-ATCGTGCGTGACATTAAGGAGAAG-3′ and reverse 5′-AGGAAGAAGGCTGGAAGAGTG-3′.

Following treatment with or without BA (50 µM) for 24 h at 37°C, pancreatic cancer cells (1×10⁶) were washed with PBS and fixed in 4% paraformaldehyde for 30 min at room temperature. Fixed cells were washed again with PBS and permeabilized in 0.5% Triton X-100 diluted in PBS for 10 min; cells on the slides were subsequently blocked with 5% bovine serum albumin (Beyotime Institute of Biotechnology, Guangzhou, China) for 1 h at room temperature and were then incubated with a Sox2 antibody (1:150 dilution) at 4°C overnight. Following incubation with a secondary antibody (1:150 dilution; cat. no. 913921; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 1 h at room temperature, the nuclei of the cells were stained with DAPI and the cells were sealed on glass slides. A confocal microscope (Zeiss AG, Oberkochen, Germany) was used to capture images and record the cells at a wavelength of 488 nm and a magnification of ×400.

Following treatment with various concentrations of BA (0, 25 and 50 µM), with BA (50 µM) for various durations (0, 5, 15, 30, 60 and 120 min), various concentrations of AICAR (0 and 2 mM), various concentrations of gemcitabine (0, 1 and 5 µM), or with BA (50 µM) combined with siRNA to knockdown AMPK or siControl, or with gemcitabine (5 µM) combined with BA (50 µM) for 24 h at 37°C, total proteins were extracted from Mia PaCa-2 and Panc-1 cells (1×10⁶) grown under experimental conditions using radioimmunoprecipitation assay lysis buffer (Beyotime Institute of Biotechnology). The bicinchoninic acid protein assay kit (Pierce; Thermo Fisher Scientific, Inc.) was used to determine protein concentrations, according to the manufacturer’s protocol. Western blotting was conducted as previously described (40). Briefly, proteins (150 µg) were separated by 10% SDS-PAGE and were transferred onto polyvinylidene difluoride membranes. Subsequently, the membranes were blocked with 5% non-fat dry milk in Tris-buffered saline containing 0.1% Tween (TBST) and were incubated with primary antibodies overnight at 4°C. After three washes in TBST (10 min/wash), the membranes were incubated with goat anti-rabbit immunoglobulin G (IgG)-horseradish peroxidase (HRP) (1:10,000 dilution; cat. no. ab6721; Abcam) or goat anti-mouse IgG-HRP (1:10,000 dilution; cat. no. ab6789; Abcam) secondary antibodies for 1 h at room temperature, and were washed again. The immunoreactive bands were visualized using a chemiluminescence detection system (EMD Millipore) through the peroxidase reaction, and images of the bands were recorded using the ChemiDoc XRS imaging system (Bio-Rad Laboratories, Inc.). β-actin was used as an internal loading control.

The present study was approved by the Ethical Committee of The First Affiliated Hospital of Xi’an Jiaotong University (Xi’an, China), and written informed consent was obtained from the patients. A total of 72 pancreatic cancer tissues and eight normal pancreatic samples were collected from the Department of Hepatobiliary Surgery, The First Affiliated Hospital of Xi’an Jiaotong University. The details of these samples were provided in our previous study (41). The tissue samples were fixed in 10% paraformaldehyde and were embedded in paraffin for further immunohistochemical analysis, as previously described (42). Briefly, pancreatic tissues were cut into 5-µm sections on glass slides, and the sections were deparaffinized and rehydrated, after which, the sections underwent antigen retrieval and endogenous enzyme blocking. The sections were then incubated with primary antibodies (anti-P-AMPK and anti-Sox2; 1:150 dilution) overnight at 4°C, followed by incubation with streptavidin peroxidase (Dako LSAB + HRP kit; Dako; Agilent Technologies, Inc., Santa Clara, CA, USA) for 30 min at room temperature. The sections were incubated with DAB for 5 min followed by counterstaining with hematoxylin for 2 min at room temperature. Finally, the sections were observed under a light microscope (Nikon Corporation).

Loss-of-function analysis was conducted using siRNA to knockdown AMPK, which was purchased from Shanghai GenePharma Co., Ltd. (Shanghai, China). The sequences of siRNA and corresponding siControl are provided in our previous study (41). Each siRNA (100 nM) was mixed with Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) as a carrier and were transfected into pancreatic cancer cells (1×10⁶) for 10 h at 37°C, according to the manufacturer’s protocol. The efficacy of AMPK knockdown was validated by western blot analysis. A total of 48 h post-transfection, these cells underwent further experimentation (Transwell invasion assay and tumorsphere formation assay).

Data shown are representative of three independent experiments, and for each experiment, at least three samples were analyzed for each treatment group. Data are presented as the means ± standard deviation. Differences were evaluated using Student’s t-test, or one-way analysis of variance for multiple comparisons with the Student-Newman-Keuls method as a post hoc test. Statistical analysis of the human tissue data was performed using Pearson’s χ² test. All statistical analyses were performed using SPSS 20.0 (SPSS, Inc., Chicago, IL, USA). P<0.05 was considered to indicate a statistically significant difference.

To determine whether BA could suppress the viability of pancreatic cancer cells, the MTT assay was used. Briefly, Mia PaCa-2 and Panc-1 cells were treated with a series of gradually increasing concentrations of BA [0 (control), 12.5, 25, 50, 100 and 200 µM] for 24, 48 and 72 h, and absorbance was measured at 490 nm at the designated time-points to analyze cell viability. The results indicated that BA impaired the growth of Mia PaCa-2 and Panc-1 cells in time- and dose-dependent manners (Fig. 1A and B). The half maximal inhibitory concentration of BA was ~50 µM for both Panc-1 and Mia PaCa-2 cells, and it exerted no cytotoxic effects on a pancreatic duct epithelial cell line (data not shown). This concentration of BA was in accordance with previously used dosages (34); therefore, 50 µM BA was used for subsequent experiments. The present study subsequently examined the effects of BA on colony formation and tumorsphere formation in Mia PaCa-2 and Panc-1 cells (Fig. 1C-F). As shown in Fig. 1C and E, compared with in untreated control cells, the number of colonies in cells treated with 50 µM BA was markedly decreased. Since cancer stemness is associated with mechanisms underlying cancer metastasis and occurrence, the present study further explored the role of BA in the stemness of pancreatic cancer. As shown in Fig. 1D and F, tumorsphere formation was decreased in Mia PaCa-2 and Panc-1 cells upon BA treatment compared with in untreated cancer cells.

It has previously been reported that the EMT process endows cancer cells with an increased self-renewal capability and mesenchymal phenotype, which is necessary for tumor metastasis (43). Therefore the present study examined the effects of BA on EMT, migration and invasion of pancreatic cancer cells. Notably, BA-treated Mia PaCa-2 and Panc-1 cells exhibited reduced migration and invasion in vitro compared with untreated cells (Fig. 2A-D). In addition, the results of western blotting suggested that the epithelial marker E-cadherin was elevated, whereas the mesenchymal marker vimentin was decreased in BA-treated Mia PaCa-2 and Panc-1 cells compared with the control (Fig. 2E and F). Collectively, these data indicated that BA may inhibit the migration and invasion of pancreatic cancer cells through targeting EMT.

Although the number of known molecular markers of the cancer stem-like phenotype is still increasing, three transcription factors; namely Sox2, Oct4 and Nanog, have been strongly validated as master regulators in the maintenance of the cancer stem-like phenotype (8). The present study examined whether BA treatment affects the expression of these master mediators of pluripotency (Fig. 3). RT-qPCR analysis revealed that BA treatment suppressed the mRNA expression levels of Sox2, Oct4 and Nanog (Fig. 3A and B). In addition, as shown in Fig. 3C, western blotting indicated that compared with the group without BA treatment, treatment of Mia PaCa-2 and Panc-1 cells with 50 µM BA downregulated the expression levels of Sox2, Oct4 and Nanog, and upregulated P-AMPK expression. The suppressive effects of BA on other stemness markers (CD133, ALDH1 and EpCAM) were similar to those on the aforementioned three master regulators (data not shown). The findings of immunofluorescence analysis also confirmed that BA could suppress the expression and nuclear localization of Sox2 in both pancreatic cancer cell lines (Fig. 3F and G). Notably, AMPK signaling has an important influence on the cancer stem-like phenotype (20,44). Therefore, to investigate whether BA could activate AMPK signaling, Mia PaCa-2 and Panc-1 cells were treated with BA at various time-points (0, 15, 30, 60, 90 and 120 min), and total protein was extracted a to detect the expression levels of P-AMPK by western blotting. The results revealed that BA treatment enhanced the expression of P-AMPK in a time-dependent manner (Fig. 3D and E). Collectively, these data indicated that BA treatment downregulated the expression of Sox2, Oct4 and Nanog, and activated AMPK signaling in pancreatic cancer cells.

To uncover the potential association between AMPK signaling and cancer stemness, immunohistochemical analysis was conducted to detect the expression levels of P-AMPK and Sox2 in normal pancreatic tissues and pancreatic cancer tissues. Immunohistochemical analysis revealed that, compared with in normal pancreatic tissues, the expression levels of P-AMPK were decreased in pancreatic cancer tissues. Conversely, the expression levels of Sox2 were markedly elevated (Fig. 4A). Furthermore, Sox2 expression was detected in 78.3% (47/60) of pancreatic cancer tissues with negative P-AMPK expression (Fig. 4B), thus suggesting an inverse relationship between P-AMPK signaling and the cancer stem-like phenotype, and indicating that downregulation of P-AMPK may lead to increased cancer stemness. In order to further clarify the effects of AMPK signaling on the maintenance of pancreatic cancer cell stemness and to determine its role in EMT, an AMPK activator, AICAR (2 mM), was used to treat Mia PaCa-2 and Panc-1 cells for 24 h. The results suggested that AMPK activation by AICAR effectively inhibited the invasion and tumorsphere formation of Mia PaCa-2 and Panc-1 cells compared with the vehicle control (Fig. 4C-F). Furthermore, the results of western blotting indicated that the expression levels of the epithelial marker E-cadherin were elevated, whereas those of the mesenchymal marker vimentin were decreased in AICAR-treated Mia PaCa-2 and Panc-1 cells compared with the control. In addition, the expression levels of the master pluripotency inducers (Sox2, Oct4 and Nanog) were decreased (Fig. 4G and H). Similar effects were observed on other stemness markers (CD133, ALDH1 and EpCAM; data not shown). These data indicated that AMPK activation by AICAR may exert similar effects to BA on EMT and stemness of pancreatic cancer cells.

In order to test the hypothesis that BA inhibits stemness and EMT of pancreatic cancer through activating AMPK signaling, AMPK-specific siRNA (si-AMPK) was used to silence AMPK expression in pancreatic cancer cells in conjunction with BA treatment. Successful silencing of AMPK expression was confirmed in Mia PaCa-2 and Panc-1 cells by western blotting (Fig. 5A). Subsequently, AMPK-depleted and control pancreatic cancer cells (Mia PaCa-2-siAMPK, Mia PaCa-2-siControl, Panc-1-siAMPK and Panc-1-siControl) were treated with BA and were subjected to Transwell invasion assay, tumorsphere formation assay and western blot analysis (Fig. 5). BA could effectively inhibit invasion (Fig. 5D-F) and tumorsphere formation (Fig. 5G-I) in Mia PaCa-2-siControl and Panc-1-siControl cells. Conversely, Mia PaCa-2-siAMPK and Panc-1-siAMPK cells exhibited a slightly higher capacity for invasion and tumor-sphere formation, which was not suppressed by BA treatment. Furthermore, this study analyzed whether BA could abrogate the expression of master pluripotency inducers (Sox2, Oct4 and Nanog) and EMT markers (E-cadherin and vimentin) in the absence of AMPK. As shown in Fig. 5B and C, BA treatment not only suppressed the expression of Sox2, Oct4, Nanog and vimentin, but also increased the expression of E-cadherin in Mia PaCa-2-siControl and Panc-1-siControl cells; however, no alterations were detected in Mia PaCa-2-siAMPK and Panc-1-siAMPK cells. Similar results were also determined with regards to other stemness markers (CD133, ALDH1 and EpCAM; data not shown). Collectively, these results indicated that BA may inhibit invasion and tumorsphere formation of pancreatic cancer cells, and regulate the expression levels of Sox2, Oct4, Nanog and EMT markers in an AMPK-dependent manner.

Gemcitabine is the major chemotherapeutic agent currently used in the treatment of pancreatic cancer; however, chemoresistance is a serious issue that markedly affects the prognosis of patients. A previous study demonstrated that gemcitabine treatment induces stemness in pancreatic cancer cells (45), which may serve a critical role in resistance to gemcitabine. The present results revealed that gemcitabine treatment dose-dependently increased the expression levels of CSC markers, including Sox2, Oct4 and Nanog (Fig. 6A and B). Similar effects were observed on other stemness markers (CD133, ALDH1 and EpCAM; data not shown). Since BA markedly inhibited the stemness of cancer cells, this study further investigated whether it could attenuate gemcitabine-induced stemness and facilitate the antitumor effects of gemcitabine. Notably, BA reversed gemcitabine-induced stemness, as revealed by decreased Sox2, Oct4 and Nanog expression (Fig. 6C and D). Since Panc-1 cells are resistant to gemcitabine (46), Panc-1 cells were selected for further experiments. Notably, BA combined with gemcitabine was more effective at suppressing proliferation (Fig. 6E) and inducing apoptosis (Fig. 6F and G) of Panc-1 cells. Taken together, these data suggested that BA may be considered a sensitizer in gemcitabine treatment.